We have added glucose and nutrients to manipulate soil microbial activity and nutrient availability in a boreal spruce forest to study the performance of birch and spruce seedlings in relation to the soil microbial community. The proportion of aboveground biomass in the seedlings was largest in plots amended with extra nutrients, while ectomycorrhizal (ECM) colonisation was low in these plots. ECM appeared beneficial for growth of both species, but only at low levels of colonisation (<25% ECM colonised root-tips). The soil microbial biomass, as determined by total PLFA, was largest in plots treated with glucose and there was a significant negative relationship between birch seedling size and levels of total PLFA in soil. This could be taken to suggest that poor seedling growth was due to nutrient limitation caused by microbial assimilation. However, the treatment response of the birch seedlings was generally weak, and spruce often showed no response at all to the addition of nutrients and glucose. The most consistent parameter for the variation in plant performance, as well as for the microbial soil community, was the block-effect. This suggests a strong spatial structure in the soil microbial community, and that this structure was robust with respect to our treatments even though they continued over a 3-year period.